Jurnal Teknologi AN IMPROVED PERTURBATION AND OBSERVATION BASED MAXIMUM POWER POINT TRACKING METHOD FOR PHOTOVOLTAIC SYSTEMS.

Transcription

1 Jurnal Teknologi AN IMPROVED PERTURBATION AND OBSERVATION BASED MAXIMUM POWER POINT TRACKING METHOD FOR PHOTOVOLTAIC SYSTEMS Ammar Hussein Mutlag a,c*, Azah Mohamed a, Hussain Shareef b a Department of Electrical, Electronic and Systems Engineering, Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, Bangi, Selangor, Malaysia b No. 1, Lorong Ayer Hitam Kawasan Institusi Penyelidikan, Kajang, 4, Kajang, Selangor, Malaysia c College of Electrical and Electronic Engineering Techniques, Middle Technical University, Baghdad-Iraq Full Paper Article history Received 23 June 2015 Received in revised form 15 November 2015 Accepted 23 January 2016 *Corresponding author ammarhussien@ukm.edu.my Graphical abstract Abstract Isc I MPP PV current (A) V MPP PV voltage (V) Voc P MPP PV power (W) In photovoltaic (PV) system, maximum power tracking (MPPT) is crucial to improve the system performance. Irradiance and temperature are the two important parameters that affect MPPT. The conventional perturbation and observation () based MPPT algorithm does not accurately track the PV maximum power point. Therefore, this paper presents an improved algorithm () based on variable perturbation. The idea behind the algorithm is to produce variable step changes in the reference current/voltage for fast tracking of the PV maximum power point. The based MPPT is designed for the 25 SolarTIFSTF-120P6 PV panels, with a capacity of 3 kw peak. A complete PV system is modeled using the MATLAB/Simulink. Simulation results showed that the based MPPT achieved faster and accurate performance compared with the conventional algorithm. Keywords: Perturb and observe algorithm; maximum power tracking; photovoltaic 2016 Penerbit UTM Press. All rights reserved 1.0 INTRODUCTION Renewable energy (RE) is presently getting more attention all around the world, particularly due to depletion of fossil fuels [1]. It is considered as the future energy source because it is clean, secure, and environmentally friendly. Among all the RE sources, photovoltaic (PV) energy systems seems to be widely applied because of the availability of enormous energy received from the sun [2]. However, PV systems have the problem of discontinuous power generation under different weather conditions [3]. In addition, the extracted power from PV system depends highly on the power-voltage (P-V) and current-voltage characteristic which vary with irradiance (G) and temperature (T) [4]. Therefore, to increase the PV system efficiency, it is crucial to operate the system at the maximum power point (MPP) which is a unique point on the P-V curve. In addition, the maximum power point tracking (MPPT) algorithm needs to be developed to increase the efficiency of PV systems [5]. Many MPPT algorithms have been mentioned in the literature which varies from simple algorithms, such as perturb and observe () [6], hill climbing (HC) [7] and incremental conductance (IC) [8], to 78: 6 2 (2016) eissn

2 20 Ammar Hussein, Azah Mohamed & Hussain Shareef / Jurnal Teknologi (Sciences & Engineering) 78: 6 2 (2016) complicated algorithms, such as fuzzy logic controller based MPPT [9], artificial neural network based MPPT [10], and other soft computing (SC) based MPPT [11]. The algorithm is one of the first algorithms that deal with MPPT issue. It is used to observe the change in power in the system. However, the algorithm does not have the ability to track accurate MPP because it suffers from oscillation around the MPP. Moreover, the method depends on the fixed step size of the current/voltage which limits its performance [12]. Another MPPT algorithm is HC which is similar to but the difference between them is that the HC method perturbs the duty cycle instead of the current/voltage. In addition, the HC approach is prone to failure in cases of large changes in weather conditions. IC is another simple MPPT algorithm which tracks the PV module power against the voltage curve to determine the MPP. However, the IC algorithm suffers from its inability to extract maximum power from the PV panel due to oscillation around the MPP. Recently, another MPPT algorithms based on artificial intelligence have been presented. Fuzzy logic controller (FLC) has been utilized for tracking MPP because it is robust and less depends on the mathematical model [13]. However, FLC depends highly on the membership functions and control rules which usually are obtained by time consuming trial and error procedure [14]. Meanwhile, ANN has been used to track the MPP. However, ANN requires large number of data for training [15]. SC optimization based MPPT algorithms have also been mentioned in the literature [16]. The disadvantage of the SC methods is that it suffers from trapping in the local minimum solutions. Therefore, there is still a need to develop a simple, fast, and accurate MPPT algorithm. This paper presents a robust, simple, fast, and accurate MPPT algorithm by improving the algorithm. 2.0 PV MODELLING The solar irradiation (G) and the temperature (T) are the main two parameters responsible for determining the operating point of PV panel and hence the MPP. The equivalent electrical circuit for the PV shown in Figure 1 is used to obtain the characteristics of a PV cell. It consists of a current source connected in parallel with resistor and diode, and a series resistor. The mathematical model of the circuit which represents the cell output current, I, is expressed as [9]: I = I ph I o (e (q(v +I.R s n.k B.T ) 1) V + I + R s (1) R sh where I is cell output current (A), Iph is light-generated current (A), Io is cell reverse saturation current or dark current (A), q is electronic charge (1.6 *10-19 C), V is cell output voltage (V), n is ideality factor, KB is Boltzmann s constant (1.38*10-23 J/K) and T is cell temperature (K). In this study, 25 SolarTIFSTF-120P6 PV modules are used to supply a 3 kw peak. The modules are arranged in series-connection configuration which produces DC output voltage of 435 V. The characteristic of a typical PV module is depicted in Table 1. G T IPh ID D PV Cell Rsh Rs Figure 1 Electrical equivalent circuit of PV cell PV module Maximum Power (PMPP) Table 1 PV module characteristics Open circuit voltage (Voc) Short circuit current (Isc) Voltage at maximum power (VMPP) Current At maximum power (IMPP) Current coefficient (α) Voltage coefficient (β) temperature temperature + V - I SolarTIFSTF-120P6 120W 21.5 V 7.63 A 17.4 V 6.89 A ma/ º C V/ º C 3.0 CONVENTIONAL BASED MPPT Load The conventional algorithm always measures the power (P) in order to find the direction of the progress which has the relationship with the reference current/voltage. The goal is to track the MPP as depicted in Figure 2. The conventional algorithm measures the current and voltage at a sampling time, t, then calculates P(t) and compares it with the previous sample P(t-1). The algorithm then continues increasing the reference current/voltage by a fixed value defined by the ratio of small change of voltage to current (Δv/Δi). If the result of the comparison [P(t)- P(t-1)] is greater than zero, it indicates that the algorithm move towards the MPP. On the hand, it will

3 21 Ammar Hussein, Azah Mohamed & Hussain Shareef / Jurnal Teknologi (Sciences & Engineering) 78: 6 2 (2016) continue decreasing the reference current/voltage by a fixed value (Δv/Δi), if the result of the comparison [P(t)- P(t-1)] is less than zero, thus, indicating that the algorithm is moving away from the MPP. The conventional algorithm does not adjust current/voltage, if the result of [P(t)- P(t-1)] is equal to zero. The performance of the conventional algorithm depends highly on the fixed step of Δv/Δi. Increasing the value of the fixed step makes the algorithm response fast but at the same time lead to large oscillation around the MPP. On the other hand, decreasing the value of the fixed step makes the response slow but leads to small oscillation around the MPP. Therefore, the conventional algorithm still needs to be improved. around MPP occurs due to the fixed Δv/Δi value. The reference current/voltage in the proposedim- algorithm can be expressed as: I, V = I, V + Ф f(x) (2) where I, V is the reference current/voltage. f(x) is one of the mathematical functions shown in Figure 3. Three functions have been investigated as shown in Figure 3. The linear equation is found to be the best function to be used in algorithm. The overall procedure of the algorithm is described in Figure 4. Δi/Δv Isc I MPP P MPP X2 PV current (A) PV power (W) Δi/Δv f(x) X1 ΔP PV voltage (V) V MPP Voc X2 Figure 2 P-V curve of a PV module 4.0 IMPROVED BASED MPPT As noted in the conventional based MPPT procedure, the main drawback of the conventional algorithm is the selection of the fixed step value of Δv/Δi. An improper selection of this fixed step value leads to poor performance of the overall system. Therefore, this paper presents an improved (Im- ) method which uses a variable step size. The variable step size means that the value of the step size Δv/Δi should be directly proportional with the value of the difference in power [P(t)-P(t-1)]. The mechanism of the depends on the following logic: - If the [P(t)-P(t-1)] >>0 then the Δv/Δi = very large - If the [P(t)-P(t-1)] >0 then the Δv/Δi= large - If the [P(t)-P(t-1)] =0 then the Δv/Δi= 0 According to this logic, the method can move quickly towards the MPP if it is far, and steps slowly if it is near the MPP, and finally stops at the MPP. In this way, the proposed algorithm is expected to achieve faster response and stability at the MPP, because the oscillation in the conventional X2 f(x) Δi/Δv f(x) X1 X1 Figure 3 Investigated equations for f(x) ΔP ΔP

4 22 Ammar Hussein, Azah Mohamed & Hussain Shareef / Jurnal Teknologi (Sciences & Engineering) 78: 6 2 (2016) Start Initialize the upper limit for the reference current,voltage Initialize the lower limit for the reference current,voltage Measure i and v to calculate p(t) behavior of the MPPT by the conventional algorithm has a large oscillation around the MPP as shown in Figure 6. Hence, the results indicated that the proposed algorithm is robust compared with the conventional algorithm No Is P(t)-p(t-1)<0? Yes Complement the slop sign Calculate the reference current,voltage using (2) No Is current, voltage >reference current Yes current, voltage =upper limit Figure 5 Speed response of the and conventional No Is current, voltage <reference current Yes 2998 current, voltage =lower limit P(t-1) = p(t) End Figure 4 Flowchart of the proposed algorithm 5.0 RESULTS AND DISCUSSION The response and performance of the MPPT to track MPP for 3 kw PV system with SolarTIFSTF-120P6 PV modules has been validated using MATLAB/Simulink. The performance of the algorithm is compared with the conventional algorithm to exhibit its capability to track the MPP under nominal condition for SolarTIFSTF-120P6 PV modules as shown in Figure 5. The figure clearly shows that the algorithm can track the 3kW power and achieve very fast response compared with the conventional algorithm. The rise time achieved by the is approximately 0.02 s meanwhile the rise time achieved by the conventional is approximately 0.57 s. Therefore, the speed response of MPPT has been significantly improved by the proposed algorithm compared to the conventional algorithm. The behavior of the MPPT based on the Im- algorithm is also characterized by a stable and oscillation free power around the MPP, meanwhile the Figure 6 Steady state responses of the and conventional For further evaluation, simulations were carried out under various solar irradiances (G) with constant temperature (T) where the irradiance is changed with a ramp from W/m 2 to 950 W/m 2 as shown in Figure 7. For this case, the response for the ramp irradiance changes as shown in Figure 8. It clearly shows that the proposed algorithm can extract more power compared with the conventional algorithm. Moreover, the achieved faster response with small oscillation meanwhile the conventional algorithm shows a large oscillation during the ramp irradiance change. The step irradiance change is also implemented for further evaluation. A step change of G from

5 23 Ammar Hussein, Azah Mohamed & Hussain Shareef / Jurnal Teknologi (Sciences & Engineering) 78: 6 2 (2016) W/m 2 to 950 W/m 2 as shown in Figure 9 was simulated. The response for the step G change is shown in Figure 10. It can be seen from Figure 10 that the proposed Im- algorithm can achieve very fast response and track the MPP with very short time compared with the conventional algorithm. Moreover, the proposed achieved the MPP with very small oscillation unlike the conventional algorithm as shown in Figure Solar irradiance (W/m 2 ) Solar irradiance (W/m 2 ) Figure 7 Ramp irradiance change Figure 9 Step irradiance change Figure 8 Response of and for the ramp irradiance change Figure 10 Response of and for the step irradiance change Another evaluation is made by carrying out simulations under various temperatures (T) with constant irradiance (G) where the T ramps from 43.6 o C to 53.6 o C as shown in Figure 11. The responses of the ramping T obtained by the and algorithms are shown in Figure 12. The extracted power from the algorithm is again greater compared with the extracted power from the conventional algorithm with much faster response time as can be seen in Figure 12.

6 24 Ammar Hussein, Azah Mohamed & Hussain Shareef / Jurnal Teknologi (Sciences & Engineering) 78: 6 2 (2016) Temperature ( o C) Figure 11 Ramp temperature change Figure 12 Response of and for the ramp temperature change The last test case is simulated with a step T change as shown in Figure 13 from 43.6 o C to 53.6 o C and the corresponding response is shown in Figure 14. The response shows that the algorithm achieves better performance compared with the conventional algorithm. The algorithm tracks the MPP with a shorter time with acceptable oscillation as can be seen in Figure 14. Temperature ( o C) Figure 13 Step temperature change Figure 14 Response of and for the step temperature change 6.0 CONCLUSION An improved algorithm based MPPT for PV systems has been proposed by considering variable step change. To validate the performance of the proposed MPPT algorithm, PV modelling with the MPPT was developed in the Matlab/Simulink environment to simulate various conditions and changes in solar irradiance and temperature. Simulation result showed that the proposed algorithm can achieve better performance compared with conventional algorithm in all conditions. The algorithm succeeds to track the MPP in all the test cases. Furthermore, the results showed that the proposed Im- algorithm is robust, simpleand accurate compared with the conventional algorithm. Acknowledgement This research work is funded by the Universiti Kebangsaan Malaysia (UKM) under project ETP References [1] Abdmouleh, Z. Alammari, R. A. M. Gastli, A Review Of Policies Encouraging Renewable Energy Integration & Best Practices. Renewable and Sustainable Energy Reviews. 45: [2] Daud, M. Z. Mohamed, A. Ibrahim, A. A. and Hannan, M.A Heuristic Optimization Of State-Of-Charge Feedback Controller Parameters For Output Power Dispatch Of Hybrid Photovoltaic/Battery Energy Storage System. Measurement. 49: [3] Mat Su, A. S. Abd Ghani R. Slamet Modelling and Simulation of Boost Converter with Maximum Power Point Tracking (MPPT) for Photovoltaic Application. Jurnal Teknologi. 71(5): 1 4 [4] Esram, T. and Chapman, P.L Comparison Of Photovoltaic Array Maximum Power Point Tracking Techniques. IEEE Transactions on Energy Conversion. 22:

7 25 Ammar Hussein, Azah Mohamed & Hussain Shareef / Jurnal Teknologi (Sciences & Engineering) 78: 6 2 (2016) [5] Zhou, T. and Sun, W Study On Maximum Power Point Tracking Of Photovoltaic Array In Irregular Shadow. Electrical Power & Energy Systems. 66: [6] Mamarelis, E. Petrone, G. and Spagnuolo, G A Two- Steps Algorithm Improving The Steady State MPPT efficiency. Applied Energy. 113: [7] Kjaer, S. B Evaluation of the Hill Climbing and the Incremental Conductance Maximum Power Point Tracking for Photovoltaic Power Systems. IEEE Transactions on Energy Conversion. 17: [8] Tey, K. S. and Mekhilef, S Modified Incremental Conductance MPPT Algorithm To Mitigate Inaccurate Responses Under Fast-Changing Solar Irradiation Level. Solar Energy. 101: [9] Altin, N. and Ozdemir, S Three-Phase Three-Level Grid Interactive Inverter With Fuzzy Logic Based Maximum Power Point Tracking Controller. Energy Conversion and Management. 69: [10] Liu, Y. H. Liu, C. L. Huang, J.W. and Chen, J. H Neural- Network-Based Maximum Power Point Tracking Methods For Photovoltaic Systems Operating Under Fast Changing Environments. Solar Energy. 89: [11] Letting,L. K. Munda, J. L. and Hamam, Y Optimization Of A Fuzzy Logic Controller For PV Grid Inverter Control Using S-Function Based PSO. Solar Energy. 86: [12] Sang-Hoon, P. Gil-Ro, C. Yong-Chae, J. and Chung-Yuen, W Design and application for PV generation system using a soft-switching boost converter with SARC. IEEE Transactions on Industrial Electronics. 57: s [13] Ali, J. A. Hannan, M.A. and Mohamed, A Rule-based Fuzzy and V/f Control for Induction Motor Speed Responses Using SVPWM Switching Technique. Przegląd Elektrotechniczny. 91(30): [14] Mutlag, A. H. Shareef, H. Mohamed, A. Hannan, M. A. and Ali, J. A An Improved Fuzzy Logic Controller Design for PV Inverters Utilizing Differential Search Optimization. International Journal of Photoenergy. 1(1): [15] Subiyanto, Mohamed, A. and Shareef H Hopfield Neural Network Optimized Fuzzy Logic Controller for MaximumPower Point Tracking in a Photovoltaic System. International Journal of Photoenergy. 1(1): [16] Benyoucef, A. Chouder, A. Kara, K. and Silvestre, S Artificial Bee Colony Based Algorithm For Maximum Power Point Tracking (MPPT) for PV systems operating under Partial Shaded Conditions. Applied Soft Computing. 32: